Making SEO A/B Testing Easy

At SearchPilot, we are on a mission to prove the value of SEO for the world’s biggest websites. It can be incredibly hard to connect specific changes to their associated SEO benefit without controlled testing. Our platform makes SEO A/B testing easy in three main ways:

  1. By automatically splitting site sections into statistically-similar groups of pages, including or excluding any groups of pages we want
  2. By making it easy to make the changes you want to test
  3. By running advanced statistical models on your analytics data to understand the impact

When you have a winning test, you can deploy it to 100% of pages using the SearchPilot platform, or build it out yourself.

This page will focus on how SearchPilot makes testing easy. If you want to learn more about what SEO testing is and how it works, you can dive deeper into that in this blog post.

Automatically splitting site sections

Deciding which pages should be in each bucket is one of the most important, yet least understood, parts of SEO testing. It’s important to create groups of pages that have similar levels of traffic, and that are statistically similar to each other. SearchPilot’s proprietary bucketing algorithm does this automatically:

Making the changes you want to test

SearchPilot is platform independent, deploying like a Content Delivery Network (CDN) - typically “inside” your existing CDN. This allows us to make server-side changes to our customers’ websites on-the-fly, and to roll back those changes as quickly as we implemented them:

Key benefits of making changes this way:

  • Deploy changes to variant pages easily - Most content management systems (CMS) offer no way of making changes to a large group of pages (but not all pages) in a site section
  • No engineering resource required - Using SearchPilot for SEO A/B testing means you won’t need any engineering resource to run tests, and you can even use our meta-CMS functionality to implement positive changes
  • Server-side changes - By deploying into your infrastructure, SearchPilot makes changes server side that are indistinguishable from any other change you might make - both to users and to search engines
  • Expert professional services team - Our professional services team means you never have to worry about getting the most value out of the platform, or having access to quality test ideas
  • Run a testing operation at scale - We run our tests on sets of templated pages that get upwards of 1k sessions / day, and our ability to split up your website into sections means the bigger your website, the more tests you can run simultaneously.

Making changes just to the set of variant pages is one of the more complicated elements of testing without SearchPilot in place:

Running complex SEO tests

The sophistication of our platform and expertise of our team means deploying the SearchPilot platform will unlock tests for you that were not previously possible, like:

Analysing the data

We use a neural network algorithm to analyse data from your analytics for our SEO A/B tests. The sophisticated mathematics of the analysis system accounts for all external and confounding factors, such as:

  • Google Algorithm Updates
  • Seasonality
  • Competitor’s changes
  • Unrelated changes to the site
  • Advertising campaigns

After each test, your dedicated professional services consultant will provide you with a report that you can share internally with your team, and use to prove business value and prioritise key changes in your development backlog. You read more about the process of reaching statistical significance here:

The importance of SEO testing in the enterprise

There are two common patterns that make SEO testing even more important in enterprise settings:

  1. It can take a very long time to get changes implemented, so proving the value of changes, and demonstrating that value to other teams is critical (note: our meta-CMS capabilities mean you can even deploy winning tests to the whole site)
  2. Many large companies also have powerful domains with strong SEO performance, and deploying changes required by product or user-focused teams can be nerve-wracking because of their potential negative impacts on organic search performance. SEO A/B testing can be useful to sense-check that changes required by the business for other reasons don’t have a significant negative impact on SEO
For our customers that are already market leaders, their biggest wins come from never deploying losers rather than from only deploying winners. Will Critchlow, CEO SearchPilot

Get a Demo

If you're interested in a short demo, please fill in this form and one of the SearchPilot team will ping you an email.

Alternatively, if you have any other questions, feel free to drop us a line at


© 2015-2022 SearchPilot. All rights reserved. Privacy policy.